| *** tpb has joined #yosys | 00:00 | |
| *** luismarques has quit IRC | 00:04 | |
| *** luismarques has joined #yosys | 00:08 | |
| *** maartenBE has quit IRC | 00:11 | |
| *** maartenBE has joined #yosys | 00:14 | |
| *** luismarques has quit IRC | 00:17 | |
| *** luismarques has joined #yosys | 00:18 | |
| *** luismarques has quit IRC | 00:40 | |
| *** digshadow has joined #yosys | 00:47 | |
| *** luismarques has joined #yosys | 00:48 | |
| emeb_mac | awygle: Interesting - tbh I haven't paid much attn to the power requirements of the various ADCs I've used. Mostly interested in high rates & resolution and that always has a cost. | 00:50 |
|---|---|---|
| awygle | emeb_mac: sure! whereas my applications are almost all narrowband and power-constrained | 00:51 |
| awygle | different constraints, different design (despite all the SDR hype ;) | 00:51 |
| emeb_mac | awygle: for really narrowband stuff it might make more sense to go with more analog in the front end. | 00:51 |
| awygle | emeb_mac: yes, that's what i've historically done. but it would be nice to be able to play more interesting protocol games | 00:52 |
| emeb_mac | my HFRX radio has very similar back-end processing as iceradio, but uses a QSD to convert RF to I/Q analog, then uses a high-res 50kSPS ADC | 00:53 |
| emeb_mac | might be a better fit for what you want | 00:53 |
| emeb_mac | of course a QSD has some of its own problems, but those are solveable... | 00:54 |
| awygle | oh that is interesting. i've never tried a QSD before | 00:54 |
| emeb_mac | http://ebrombaugh.studionebula.com/radio/hfrx/index.html | 00:54 |
| tpb | Title: HFRX (at ebrombaugh.studionebula.com) | 00:54 |
| emeb_mac | awygle: you probably have seen the circuit used - just not always called that. | 00:55 |
| awygle | i had put together a similar design (in the sense of using a high-end STM) but with a more conventional downconversion stage, but i couldn't get it to hit all three of performance/power/price targets | 00:55 |
| emeb_mac | yeah - designing radios is a vast multi-dimensional space to optimize | 00:57 |
| awygle | yup yup | 00:57 |
| awygle | sadly i'm mostly out of this game now but i'll definitely do more reading on this approach | 00:58 |
| emeb_mac | that STM32F373 is a fun part - it's got some really nice 16-bit differential ADCs that are surprisingly low noise. | 00:58 |
| emeb_mac | worked pretty well for doing the DSP on the back end of that quadrature ZIF front-end. | 00:58 |
| awygle | it appears from my notes i was looking at the F303 | 00:58 |
| emeb_mac | Very similar part, but w/o the good ADCs | 00:59 |
| awygle | i think i might have been aiming for higher sampling rate and paying less attention to performance | 01:00 |
| awygle | but it was a long while back | 01:00 |
| emeb_mac | If power is an issue then iceRadio was definitely not optimized for that. | 01:01 |
| emeb_mac | especially with the R820T2 VHF front end - that thing gets pretty toasty. | 01:01 |
| awygle | the interesting aspect of that design from a power perspective is that the ice40up potentially represents a very low-power way to do the baseband processing | 01:03 |
| awygle | although it's sort of unclear to me how it would compare to something like an STM | 01:03 |
| awygle | (i'll freely admit i am better at the frontend than the digital stuff) | 01:04 |
| emeb_mac | Well, the FPGA is fairly low-power, but I use it to do the wideband downconversion all digitally. An STM32 doesn't have the horsepower to handle 40MSPS data in any way. | 01:05 |
| emeb_mac | That high sample rate means it's able handle most of the HF band in one gulp, but it immediately reduces it down to a ~40kHz bandwidth, so some might argue there's more efficient ways to do that. | 01:08 |
| *** sklv has quit IRC | 01:09 | |
| awygle | ideally i would like to take in something like a 500k-2Mbps QPSK signal and put demodulated bits or even higher-level packets out over something like I2S. perhaps including FEC in the fpga as well | 01:10 |
| *** luismarques has quit IRC | 01:10 | |
| awygle | ah well. someday. too many irons in the fire already to work on something like that now. | 01:11 |
| emeb_mac | That's probably doable in a UP | 01:12 |
| *** luismarques has joined #yosys | 01:18 | |
| *** luismarques has quit IRC | 01:24 | |
| *** digshadow has quit IRC | 01:42 | |
| *** luismarques has joined #yosys | 01:43 | |
| *** luismarques has quit IRC | 01:48 | |
| *** luismarques has joined #yosys | 01:49 | |
| *** seldridge has quit IRC | 02:00 | |
| *** luismarques has quit IRC | 02:15 | |
| *** luismarques has joined #yosys | 02:17 | |
| *** digshadow has joined #yosys | 02:24 | |
| *** luismarques has quit IRC | 02:34 | |
| ZipCPU | Woohoo! All six butterfly forms are now working. That means the whole FFT works as well! Yaaaaayyyy ... | 02:54 |
| ZipCPU | Must mean its time for bed. | 02:54 |
| ZipCPU | Sigh. Celebrating too soon. I had the first three multiplies working, and ... they aren't anymore. I'll have to check the git history on Monday. *SO* close. | 02:56 |
| ZipCPU | s/multiplies/butterflies/ | 03:02 |
| *** m_w has quit IRC | 05:28 | |
| *** leviathan has joined #yosys | 06:27 | |
| *** xerpi has joined #yosys | 06:40 | |
| *** xerpi has quit IRC | 06:42 | |
| *** xerpi has joined #yosys | 06:42 | |
| *** xerpi has quit IRC | 06:44 | |
| *** xerpi has joined #yosys | 06:44 | |
| *** emeb_mac has quit IRC | 07:07 | |
| *** dys has joined #yosys | 07:22 | |
| *** proteus-guy has quit IRC | 07:37 | |
| *** jwhitmore has joined #yosys | 08:20 | |
| *** digshadow has quit IRC | 09:21 | |
| *** m_t has joined #yosys | 09:28 | |
| *** m_w has joined #yosys | 09:45 | |
| *** dys has quit IRC | 09:46 | |
| *** proteus-guy has joined #yosys | 09:49 | |
| *** m_w has quit IRC | 09:51 | |
| *** xerpi has quit IRC | 10:23 | |
| *** dys has joined #yosys | 10:31 | |
| *** sklv has joined #yosys | 10:42 | |
| *** jwhitmore has quit IRC | 11:12 | |
| *** jwhitmore has joined #yosys | 11:13 | |
| *** ar3itrary has joined #yosys | 11:22 | |
| *** jwhitmore has quit IRC | 11:42 | |
| *** _whitelogger has quit IRC | 12:16 | |
| *** _whitelogger has joined #yosys | 12:18 | |
| *** proteus-guy has quit IRC | 14:05 | |
| *** emeb_mac has joined #yosys | 14:06 | |
| *** xerpi has joined #yosys | 14:28 | |
| *** proteus-guy has joined #yosys | 14:37 | |
| *** lutsabound has joined #yosys | 14:59 | |
| *** seldridge has joined #yosys | 15:19 | |
| *** dxld has quit IRC | 15:48 | |
| *** dxld has joined #yosys | 15:49 | |
| *** digshadow has joined #yosys | 16:21 | |
| *** m_w has joined #yosys | 16:24 | |
| *** m_w has quit IRC | 16:34 | |
| *** m_t has quit IRC | 16:55 | |
| *** promach_ has quit IRC | 16:59 | |
| *** seldridge has quit IRC | 17:30 | |
| *** seldridge has joined #yosys | 17:47 | |
| *** X-Scale has quit IRC | 17:54 | |
| *** xerpi has quit IRC | 17:55 | |
| *** X-Scale has joined #yosys | 18:29 | |
| *** sklv has quit IRC | 19:23 | |
| *** sklv has joined #yosys | 19:23 | |
| *** xerpi has joined #yosys | 20:31 | |
| *** Noname_Matt has joined #yosys | 21:05 | |
| *** leviathan has quit IRC | 21:21 | |
| *** Groomblecom has joined #yosys | 21:24 | |
| *** jwhitmore has joined #yosys | 21:49 | |
| *** dys has quit IRC | 21:50 | |
| *** maartenBE has quit IRC | 21:53 | |
| *** maartenBE has joined #yosys | 21:56 | |
| *** jwhitmore has quit IRC | 22:50 | |
| *** Noname_Matt has quit IRC | 22:54 | |
| *** sklv has quit IRC | 22:59 | |
| *** danieljabailey has quit IRC | 23:04 | |
| *** danieljabailey has joined #yosys | 23:04 | |
| *** seldridge has quit IRC | 23:19 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!